This repository contains a Fluent CFF (HDF5) reader implemented for the inspection, extraction, and export of mesh and results data from Ansys Fluent .cas.h5 and .dat.h5 files, which can then be used for visualization or machine learning workflows.
-
A demo based on the Ansys
Static Mixertutorial is provided as a Jupyter Notebook:FluentCFFReader_Demo.ipynb. -
A built-in memory and time logging utility (activated via
VERBOSE = 1) records detailed runtime statistics, i.e., memory usage and execution time across all major functions. -
Performance Note:
The parser is highly optimized for large-scale industrial CFD models and has been extensively tested on complex cases containing millions of degrees of freedom.
-
Mesh Inspection
- Inspect mesh structure, surfaces, and volumetric zones.
- Retrieve volumetric point clouds.
- Extract boundary faces by name or boundary condition type.
-
Results Inspection
- Explore all available physical fields in
.dat.h5files. - List cell-centered and face-centered field variables.
- Explore all available physical fields in
-
Export Capabilities
- Export boundary surfaces as
.vtpfiles. - Export volume meshes as
.vtufiles. - Export face and cell field data as
.vtpand.vtufiles respectively.
- Export boundary surfaces as
Note: Some exported
.vtufiles contain polyhedral cells generated by PyVista.
Due to known issues in the released versions of ParaView, these files may not display correctly.
It is recommended to use the current development (nightly) version of ParaView for proper visualization of all volume meshes and cell data:
ParaView Nightly Builds
pip install -r requirements.txt